Version: 1.8.0

tabsdata.tableframe.udf.function


`UDF`
`UDF.columns`
`UDF.on_batch`	Creating UDFs:
`UDF.on_element`	Creating UDFs:
`UDF.with_columns`
`UDFList`	Abstract base class for UDFs that use list-style parameter passing.
`UDFList.columns`
`UDFList.on_batch`	Creating UDFs:
`UDFList.on_element`	Creating UDFs:
`UDFList.with_columns`
`UDFUnpacked`	Abstract base class for UDFs that use unpacked-style parameter passing.
`UDFUnpacked.columns`
`UDFUnpacked.on_batch`	Creating UDFs:
`UDFUnpacked.on_element`	Creating UDFs:
`UDFUnpacked.with_columns`

UDF

classView source ↗

class UDF(
    output_columns: list[tuple[str, Boolean | Categorical | Date | Datetime | Decimal | Duration | Enum | Float16 | Float32 | Float64 | Int8 | Int16 | Int64 | Int32 | Int128 | Null | String | Time | UInt8 | UInt16 | UInt32 | UInt64]] | tuple[str, Boolean | Categorical | Date | Datetime | Decimal | Duration | Enum | Float16 | Float32 | Float64 | Int8 | Int16 | Int64 | Int32 | Int128 | Null | String | Time | UInt8 | UInt16 | UInt32 | UInt64],
)

Bases: ABC

property

signatureLiteral['list', 'unpacked']

Defines how parameters are passed to on_batch and on_element methods.

Returns: "list": Parameters are passed as a single list (default). "unpacked": Each parameter is passed as a separate argument.

Override this property in your UDF subclass to change the parameter style.

method

View source ↗

columns

def columns(
,
) -> list[tuple[str, Boolean | Categorical | Date | Datetime | Decimal | Duration | Enum | Float16 | Float32 | Float64 | Int8 | Int16 | Int64 | Int32 | Int128 | Null | String | Time | UInt8 | UInt16 | UInt32 | UInt64]]

method

View source ↗

on_batch

def on_batch(*args) -> list[Series]

Creating UDFs:

Subclass :class:tabsdata.tableframe.udf.function.UDF.
Implement __init__ to call super().__init__(output_columns) where output_columns is a tuple or list of tuples (name, data type) specifying the UDF default output schema (column names and data types). Each tuple must contain a column name (string) and a data type (DataType).
Override exactly one of on_batch or on_element, to implement the UDF function logic.
Return a list of TabsData Series (for on_batch) or TabsData supported scalars (for on_element) with the same length as specified in the output schema.
If overriding the on_batch method, the return type must be a list of TabsData Series. If overriding the on_element method, the return type must be a list of supported TabsData scalar values. For both cases, the number of elements in the returned lists must match the number of elements in the output_columns list provided to the UDF constructor.

Using UDFs:

Instantiate a function created as above.
Pass it to TableFrame method udf().
Optionally use :meth:UDF.output_columns to override output column names or data types after instantiation.
By default, on_batch receives a list of series. Override the signature property to return "unpacked" to receive each series as a separate argument instead.

method

View source ↗

on_element

def on_element(*args) -> list[Any]

Creating UDFs:

Subclass :class:tabsdata.tableframe.udf.function.UDF.
Implement __init__ to call super().__init__(output_columns) where output_columns is a tuple or list of tuples (name, data type) specifying the UDF default output schema (column names and data types). Each tuple must contain a column name (string) and a data type (DataType).
Override exactly one of on_batch or on_element, to implement the UDF function logic.
Return a list of TabsData Series (for on_batch) or TabsData supported scalars (for on_element) with the same length as specified in the output schema.
If overriding the on_batch method, the return type must be a list of TabsData Series. If overriding the on_element method, the return type must be a list of supported TabsData scalar values. For both cases, the number of elements in the returned lists must match the number of elements in the output_columns list provided to the UDF constructor.

Using UDFs:

Instantiate a function created as above.
Pass it to TableFrame method udf().
Optionally use :meth:UDF.output_columns to override output column names or data types after instantiation.
By default, on_element receives a list of values. Override the signature property to return "unpacked" to receive each value as a separate argument instead.

method

View source ↗

with_columns

def with_columns(
    output_columns: tuple[str | None, Boolean | Categorical | Date | Datetime | Decimal | Duration | Enum | Float16 | Float32 | Float64 | Int8 | Int16 | Int64 | Int32 | Int128 | Null | String | Time | UInt8 | UInt16 | UInt32 | UInt64 | NoneType] | list[tuple[str | None, Boolean | Categorical | Date | Datetime | Decimal | Duration | Enum | Float16 | Float32 | Float64 | Int8 | Int16 | Int64 | Int32 | Int128 | Null | String | Time | UInt8 | UInt16 | UInt32 | UInt64 | NoneType]] | dict[int, tuple[str | None, Boolean | Categorical | Date | Datetime | Decimal | Duration | Enum | Float16 | Float32 | Float64 | Int8 | Int16 | Int64 | Int32 | Int128 | Null | String | Time | UInt8 | UInt16 | UInt32 | UInt64 | NoneType]],
) -> UDF

UDFList

classView source ↗

class UDFList(output_columns)

Bases: UDF, ABC

Abstract base class for UDFs that use list-style parameter passing.

When subclassing UDFList, implement on_batch or on_element with list signature:

on_batch(self, series: list[Series]) -> list[Series]
on_element(self, values: list[Any]) -> list[Any]

property

signatureLiteral['list', 'unpacked']

Defines how parameters are passed to on_batch and on_element methods.

Returns: "list": Parameters are passed as a single list (default). "unpacked": Each parameter is passed as a separate argument.

Override this property in your UDF subclass to change the parameter style.

method

View source ↗

columns

def columns(
,
) -> list[tuple[str, Boolean | Categorical | Date | Datetime | Decimal | Duration | Enum | Float16 | Float32 | Float64 | Int8 | Int16 | Int64 | Int32 | Int128 | Null | String | Time | UInt8 | UInt16 | UInt32 | UInt64]]

method

View source ↗

on_batch

def on_batch(*args) -> list[Series]

Creating UDFs:

Subclass :class:tabsdata.tableframe.udf.function.UDF.
Implement __init__ to call super().__init__(output_columns) where output_columns is a tuple or list of tuples (name, data type) specifying the UDF default output schema (column names and data types). Each tuple must contain a column name (string) and a data type (DataType).
Override exactly one of on_batch or on_element, to implement the UDF function logic.
Return a list of TabsData Series (for on_batch) or TabsData supported scalars (for on_element) with the same length as specified in the output schema.
If overriding the on_batch method, the return type must be a list of TabsData Series. If overriding the on_element method, the return type must be a list of supported TabsData scalar values. For both cases, the number of elements in the returned lists must match the number of elements in the output_columns list provided to the UDF constructor.

Using UDFs:

Instantiate a function created as above.
Pass it to TableFrame method udf().
Optionally use :meth:UDF.output_columns to override output column names or data types after instantiation.
By default, on_batch receives a list of series. Override the signature property to return "unpacked" to receive each series as a separate argument instead.

method

View source ↗

on_element

def on_element(*args) -> list[Any]

Creating UDFs:

Subclass :class:tabsdata.tableframe.udf.function.UDF.
Implement __init__ to call super().__init__(output_columns) where output_columns is a tuple or list of tuples (name, data type) specifying the UDF default output schema (column names and data types). Each tuple must contain a column name (string) and a data type (DataType).
Override exactly one of on_batch or on_element, to implement the UDF function logic.
Return a list of TabsData Series (for on_batch) or TabsData supported scalars (for on_element) with the same length as specified in the output schema.
If overriding the on_batch method, the return type must be a list of TabsData Series. If overriding the on_element method, the return type must be a list of supported TabsData scalar values. For both cases, the number of elements in the returned lists must match the number of elements in the output_columns list provided to the UDF constructor.

Using UDFs:

Instantiate a function created as above.
Pass it to TableFrame method udf().
Optionally use :meth:UDF.output_columns to override output column names or data types after instantiation.
By default, on_element receives a list of values. Override the signature property to return "unpacked" to receive each value as a separate argument instead.

method

View source ↗

with_columns

def with_columns(
    output_columns: tuple[str | None, Boolean | Categorical | Date | Datetime | Decimal | Duration | Enum | Float16 | Float32 | Float64 | Int8 | Int16 | Int64 | Int32 | Int128 | Null | String | Time | UInt8 | UInt16 | UInt32 | UInt64 | NoneType] | list[tuple[str | None, Boolean | Categorical | Date | Datetime | Decimal | Duration | Enum | Float16 | Float32 | Float64 | Int8 | Int16 | Int64 | Int32 | Int128 | Null | String | Time | UInt8 | UInt16 | UInt32 | UInt64 | NoneType]] | dict[int, tuple[str | None, Boolean | Categorical | Date | Datetime | Decimal | Duration | Enum | Float16 | Float32 | Float64 | Int8 | Int16 | Int64 | Int32 | Int128 | Null | String | Time | UInt8 | UInt16 | UInt32 | UInt64 | NoneType]],
) -> UDF

UDFUnpacked

classView source ↗

class UDFUnpacked(output_columns)

Bases: UDF, ABC

Abstract base class for UDFs that use unpacked-style parameter passing.

When subclassing UDFUnpacked, implement on_batch or on_element with unpacked signature:

on_batch(self, col1: Series, col2: Series, ...) -> list[Series]
on_element(self, val1: Any, val2: Any, ...) -> list[Any]

property

signatureLiteral['list', 'unpacked']

Defines how parameters are passed to on_batch and on_element methods.

Returns: "list": Parameters are passed as a single list (default). "unpacked": Each parameter is passed as a separate argument.

Override this property in your UDF subclass to change the parameter style.

method

View source ↗

columns

def columns(
,
) -> list[tuple[str, Boolean | Categorical | Date | Datetime | Decimal | Duration | Enum | Float16 | Float32 | Float64 | Int8 | Int16 | Int64 | Int32 | Int128 | Null | String | Time | UInt8 | UInt16 | UInt32 | UInt64]]

method

View source ↗

on_batch

def on_batch(*args) -> list[Series]

Creating UDFs:

Subclass :class:tabsdata.tableframe.udf.function.UDF.
Implement __init__ to call super().__init__(output_columns) where output_columns is a tuple or list of tuples (name, data type) specifying the UDF default output schema (column names and data types). Each tuple must contain a column name (string) and a data type (DataType).
Override exactly one of on_batch or on_element, to implement the UDF function logic.
Return a list of TabsData Series (for on_batch) or TabsData supported scalars (for on_element) with the same length as specified in the output schema.
If overriding the on_batch method, the return type must be a list of TabsData Series. If overriding the on_element method, the return type must be a list of supported TabsData scalar values. For both cases, the number of elements in the returned lists must match the number of elements in the output_columns list provided to the UDF constructor.

Using UDFs:

Instantiate a function created as above.
Pass it to TableFrame method udf().
Optionally use :meth:UDF.output_columns to override output column names or data types after instantiation.
By default, on_batch receives a list of series. Override the signature property to return "unpacked" to receive each series as a separate argument instead.

method

View source ↗

on_element

def on_element(*args) -> list[Any]

Creating UDFs:

Subclass :class:tabsdata.tableframe.udf.function.UDF.
Implement __init__ to call super().__init__(output_columns) where output_columns is a tuple or list of tuples (name, data type) specifying the UDF default output schema (column names and data types). Each tuple must contain a column name (string) and a data type (DataType).
Override exactly one of on_batch or on_element, to implement the UDF function logic.
Return a list of TabsData Series (for on_batch) or TabsData supported scalars (for on_element) with the same length as specified in the output schema.
If overriding the on_batch method, the return type must be a list of TabsData Series. If overriding the on_element method, the return type must be a list of supported TabsData scalar values. For both cases, the number of elements in the returned lists must match the number of elements in the output_columns list provided to the UDF constructor.

Using UDFs:

Instantiate a function created as above.
Pass it to TableFrame method udf().
Optionally use :meth:UDF.output_columns to override output column names or data types after instantiation.
By default, on_element receives a list of values. Override the signature property to return "unpacked" to receive each value as a separate argument instead.

method

View source ↗

with_columns

def with_columns(
    output_columns: tuple[str | None, Boolean | Categorical | Date | Datetime | Decimal | Duration | Enum | Float16 | Float32 | Float64 | Int8 | Int16 | Int64 | Int32 | Int128 | Null | String | Time | UInt8 | UInt16 | UInt32 | UInt64 | NoneType] | list[tuple[str | None, Boolean | Categorical | Date | Datetime | Decimal | Duration | Enum | Float16 | Float32 | Float64 | Int8 | Int16 | Int64 | Int32 | Int128 | Null | String | Time | UInt8 | UInt16 | UInt32 | UInt64 | NoneType]] | dict[int, tuple[str | None, Boolean | Categorical | Date | Datetime | Decimal | Duration | Enum | Float16 | Float32 | Float64 | Int8 | Int16 | Int64 | Int32 | Int128 | Null | String | Time | UInt8 | UInt16 | UInt32 | UInt64 | NoneType]],
) -> UDF